Vol. 11: 207–213, 2010 ENDANGERED RESEARCH Published online May 14 doi: 10.3354/esr00280 Endang Species Res

OPENPEN ACCESSCCESS Crowned rarely prey on in central : persecution is not justified

José Hernán Sarasola1, 3,*, Miguel Ángel Santillán1, Maximiliano Adrián Galmes1, 2

1Centro para el Estudio y Conservación de las Aves Rapaces en Argentina (CECARA), Universidad Nacional de La Pampa-CONICET, Avda. 151, 6300 Santa Rosa, La Pampa, Argentina 2The Peregrine Fund, 5668 West Flying Hawk Lane, Boise, Idaho 83709, USA

3Present address: Estación Biológica de Doñana – CSIC, Avda. Américo Vespucio s/n, 41092 Sevilla, Spain

ABSTRACT: Raptors have been reported to prey on livestock, causing considerable conflicts between of prey and local communities. Previous studies have documented that human persecu- tion is the most important threat to the endangered crowned Harpyhaliaetus coronatus in cen- tral Argentina, due to a local belief that crowned eagles heavily and consistently prey on livestock. However, there are no empirical data supporting this assertion. Such information is crucial to evalu- ating possible measures to mitigate this human–wildlife conflict. We evaluated the feeding ecology of crowned eagles in semiarid of central Argentina during the breeding seasons of 2002 to 2009. We also evaluated whether eagles are responsible for livestock and examined spa- tial-temporal variation in crowned eagles’ food habits. We identified 598 prey items consisting almost entirely of native prey: (67.7%), (16.2%), birds (3.3%), (2.5%), and inverte- brates (16.9%). We only recorded 1 (0.17%) occurrence of livestock prey remains, belonging to a domestic Capra hircus. Occurrences of the 4 main prey groups were not affected by type or season. However, reptiles were recorded in higher numbers at sites where the diet of eagles was addressed by direct observation and video recording. Contrary to other human–predator conflicts worldwide, and assuming that the single livestock prey was not scavenged, our results show that crowned eagles rarely prey on livestock. We advocate reducing human–wildlife conflicts by imple- menting management and conservation measures and by educating local communities with respect to the ecological role of crowned eagles and other predators.

KEY WORDS: Crowned eagle · Harpyhaliaetus coronatus · Conservation · Persecution · Human–wildlife conflict · Feeding ecology · Central Argentina

Resale or republication not permitted without written consent of the publisher

INTRODUCTION prey. For example, it is the main cause for the popula- tion decline of golden eagles chrysaetus in Large predators are among the most difficult species (Whitfield et al. 2004) and the principal factor to preserve: they occur at extremely low densities, are limiting population growth of hen harriers Circus cya- distributed across large areas, and may come into con- neus in the UK (Thirgood et al. 2000). Dealing with flict with human populations in parts of their geo- these conflicts requires data on the ecological circum- graphic ranges. These conflicts mainly consist of pre- stances associated with livestock predation, as well as dation on livestock (Thirgood et al. 2005), which can on the relative economic impacts of such conflicts. result in the predator being persecuted, trapped, or These data are crucial for the establishment of conser- killed, with corresponding negative effects on its con- vation plans and management decisions that aim to servation (Etheridge et al. 1997). Human persecution mitigate or avoid conflicts between wildlife and local has been identified as a serious threat to many birds of human communities.

*Email: [email protected] © Inter-Research 2010 · www.int-res.com 208 Endang Species Res 11: 207–213, 2010

The crowned eagle Harpyhaliaetus coronatus is one of sive habitat types in Argentina, the Espinal and the the largest birds of prey in , ranging from Monte biomes (Cabrera 1976), which are also southern to northern in Argentina (Fer- 2 of the representative habitat types for crowned gusson-Lees & Christie 2001). It mainly inhabits open eagles in southern South America. The Espinal in xerophytic typical of several biomes ecoregion is located in central Argentina, extending in Brazil, , and Argentina. The crowned eagle from central Santa Fe through the of Cór- has been included in the threatened fauna lists of Ar- doba and San Luis to the south of La Pampa gentina as an endangered species (López-Lanús et al. province. Vegetation types include deciduous xero- 2008, Chebez et al. 2008) and is currently considered ex- phytic , palm groves, grassy savannahs, grassy tinct in Uruguay, where it has not been reported since steppes, and bushy steppes. Climax communities are 1930 (Alvarez 1933). The International Union for the dominated by of the , repre- Conservation of Nature (IUCN) listed it as a Vulnerable sented by the caldén P. caldenia in La Pampa species until 2004, when it was placed in the Endan- province. The is located in north- gered category. The world population is estimated at less central Argentina and extends along the eastern than 1000 individuals, and populations are declining foothills of the Andes until it reaches the Patagonian (BirdLife International 2008). Despite its wide range, crit- steppe. This ecoregion is dominated by occasional ical , and ongoing population de- open scrublands formed by resinous evergreen creases, few studies exist on wild populations of this spe- bushes. In both ecoregions, the climate is temperate- cies, and many aspects of its biology are still unknown, arid with very little rainfall (between 80 and 250 mm including habitat use, habitat selection, and feeding and between 350 and 450 mm yr–1 for the Monte ecology. Although suspected threats to crowned eagles Desert and Espinal biomes, respectively; Fernandez include habitat loss (Bellocq et al. 1998) and other hu- & Busso 1997). man-related mortality (i.e. electrocution on power lines), Methods. Between 2002 and 2009, we collected only human persecution has been adequately docu- information on the food habits of crowned eagles at mented. A recent study by Sarasola & Maceda (2006), 15 breeding territories from an area where human based on interviews with local landowners and rural persecution has previously been reported (Fig. 1). workers, indicated that (1) human persecution is the Sampling was conducted during 1 of the 2 local most important cause of mortality of crowned eagles in lambing seasons in this area (spring-summer), when Central Argentina; (2) the number of eagles found killed, lambs are suspected to be more susceptible to preda- injured, or trapped was relatively high (n = 20); and (3) tion. Pellets and prey remains were collected at 11 local people cited eagle predation on livestock as the jus- nests and at perch sites used by adults near the nests. tification for persecution. Unfortunately, the food habits Pellets were hydrated and broken apart by hand, and of crowned eagles have been only poorly and anecdo- prey remains were identified using reference collec- tally documented (Maceda et al. 2003, Maceda 2007, tions at CECARA (Universidad Nacional de La Chebez et al. 2008), so little evidence exists to verify Pampa, Argentina) and identification keys (Cei 1986, whether this local belief is grounded in the reality of cur- Pearson 1995) to the lowest possible taxonomic level. rent predatory habits of these birds. We thus examined In addition, we determined eagle diets at 4 nest sites the feeding ecology of crowned eagles in an area of cen- from mid-December to mid-February by recording tral Argentina where human persecution was previously prey items that adult eagles delivered to the documented. We evaluated whether eagles are respon- nestlings. First, 2 nests located near the Jagüel del sible for livestock predation and whether livestock con- Monte area were monitored, one in 2004 and the stitutes a high percentage of prey items in the diet of other in 2005, with a time-lapse video recorder (VCR crowned eagles, or is at least similar to those reported for Panasonic AG-1070 DC). The camera was housed in other documented raptor–human conflicts. In addition, a waterproof compartment and placed approximately we examined spatial and temporal variation in the diet of 1 m above the nest. Power and video lines from the crowned eagles by evaluating the occurrence of main camera were connected to a time-lapse VCR located prey items in different habitats and seasons. on the ground about 100 m from the nest. Power was supplied to the video unit by one 12 V battery con- nected to a solar panel. The VCR was housed in a MATERIALS AND METHODS waterproof compartment and was programmed to record during two 8 h time blocks (from 06:00 to Study area. Our study was conducted in an area of 14:00 h and from 14:00 to 22:00 h). Because little is ca. 10 000 km2 in central and western parts of La known about the timing of hunting by crowned Pampa province, Argentina (Fig. 1). The natural eagles, these time blocks were selected randomly. In landscape in this region includes 2 of the most exten- 2004, the camera was placed at the nest when the Sarasola et al.: Feeding ecology and persecution of crowned eagles 209

We evaluated spatial and temporal variation of prey consumed by crown- ed eagles in western La Pampa. For each prey group, we built generalized linear models (GLMs; McCullagh & Nelder 1989) where the response vari- able was the number of individuals consumed in each of 4 prey categories (mammals, birds, reptiles, and fish), and the independent variables were habitat (2-level factor: forest or scrub- land) and year. We considered each breeding territory as a sample and in- corporated the independent variables related to the breeding sites (habitat type and year when sampling was con- ducted). Habitat features are important since they may determine local pres- ence and availability of different prey types; by including year as the inde- pendent variable, we attempted to account for temporal variability in con- sumption of each prey group. We also included a third variable related to the methodology employed to survey the food habits at each site (i.e. pellets and prey remains, direct observations, and video recording), since using different methods to study raptor diet can lead to biased representation of certain prey Fig. 1. Study area (37° 02’ S, 65° 34’ W) in La Pampa province, central Argentina, types (Margalida et al. 2007). By showing locations of crowned eagle breeding territories where diet information including this variable in the ana- s was obtained ( ) and incidents of human persecution reported by Sarasola & lysis, we evaluated possible differences Maceda (2006) (d). The thick dashed line represents the boundary between the Monte Desert (west) and Espinal (east) biomes among breeding sites linked to metho- dology-dependent biases in prey iden- tification at a coarse taxonomic level young eagle was about 1 wk old; in 2005, the camera (i.e. order). For modeling purposes, we considered only was placed when the was still unhatched. In the those sites sampled where the total number of identi- 2007 and 2008 breeding seasons, dawn-to-dusk direct fied prey was >15 (N = 13 breeding sites). All models observations were conducted at 2 nests, one in each were built using a Poisson error distribution and a log year. Observations were conducted between 06:00 link function. and 22:00 h using a 20–60× spotting scope (Bushnell) The significance of habitat, method, and year as and 2 observers located 300 m from the nest to mini- explanatory variables of consumption of each prey - mize disturbance. egory was tested using information theory. We built 8 We calculated total prey biomass in the diet by mul- models (all possible variable combinations of main tiplying the mean mass of each prey type by the num- effects without interactions) for each prey category ber of that prey recorded in the diet of crowned eagles. species and used the Akaike Information Criterion To compute prey biomass, we obtained the mean body (AIC, Burnham & Anderson 2002), which takes into mass of small species and arthropods from account both the information explained by the model Sarasola et al. (2003, 2007). J. Maceda (unpubl. data) and its complexity in terms of number of estimated provided data for the body mass of , , parameters, generating a rank from the best to the birds, and lagomorphs, while body mass of carnivores least likely model. Within a prey category, each model was obtained from Redford & Eisenberg (1992). was considered as a hypothesis explaining the con- Unidentified prey items were not considered for prey sumption of that prey. For this set of models, we first biomass calculation. calculated the second-order AIC (AICc), which is simi- 210 Endang Species Res 11: 207–213, 2010

lar to AIC but corrected for small sample size, the DISCUSSION ΔAICc (the differences in AICc with respect to the AICc of the best candidate model), and AICc weight Contrary to local beliefs, our results show that (w). The best hypothesis was weighed against the oth- crowned eagles in semiarid habitats of central ers using AICc weight, which gives an estimation of Argentina do not prey on livestock on a regular basis. the likelihood of the hypothesis given the data. Corre- In the southern Espinal and Monte desert biomes, lations between independent variables were tested, crowned eagles preyed almost exclusively on native showing non-significant correlations (p > 0.20). Statis- wild species, mainly armadillos and other small to tical analyses were performed using S-PLUS 2000 soft- medium-sized mammals and snakes. Nestling eagles ware (MathSoft 1999). were also fed natural prey items, despite the avail- ability of domestic lambs. The relative abundance of snakes in the diet was in agreement with Amadon’s RESULTS (1982) predictions regarding the feeding behavior of crowned eagles based on morphology (e.g. rough, In total, 513 prey items were identified from 67 pel- reticulate tarsi that protect against venomous lets and 221 prey remains collected during the study. bites). Dawn-to-dusk observations at nests totaled 280 h, and Only 1 domestic goat was recorded as a prey item video observations totaled 400 h. In total, 85 prey (less than 0.2% of the total items recorded), indicating delivered to young eagles by their parents were the occasional nature of such feeding behavior. This recorded during direct observations and video record- single report came from a nest in western La Pampa, ing sessions at nests. Medium-sized and small mam- where the number of eagles killed is lower than mals accounted for the majority of prey items reported for the Espinal (Fig. 1).This rate of predation recorded (edentates: 48.6%, and small mar- on livestock, assuming that this single goat was not supials: 13.7%, small carnivores and lagomorphs: scavenged as carrion, a feeding behavior previously 5.2%,), followed by (16.9%), reptiles documented for crowned eagles (Maceda et al. 2003), (16.2%), birds (3.3%), and fish (2.6%; Table 1). Rep- is considerably lower than those observed in other tiles were mainly represented by snakes, including well-documented conflicts between raptors and venomous species such as the crossed pit viper Both- worldwide. For example, domestic lamb re- rops alternatus and the Patagonian pit viper B. mains represented 1.1% to 8% of the prey items in the ammodytoides. The diet of crowned eagles was inde- diet of black eagles Aquila verreauxii in South pendent of livestock farming. We recorded livestock (Davies 1999), and a survey on lamb carcasses at 2 as a prey item (0.2%) at only 1 nest: a humerus of a ranches in Montana (USA) determined that golden domestic goat Capra hircus. Livestock prey were not eagles were responsible for an average of 24% of recorded at any of the nests monitored by direct losses between 1974 and 1985 (Matchett & observations or video recording. O’Gara 1987). Similarly, it was calculated that hen har- The consumption of mammals, fish, and birds did riers could remove up to 24% of red grouse chicks on not vary spatially or temporally, as it was indepen- grouse moors in the UK (Redpath 1991). dent of year/season and habitat type. However, fish Our results support the argument that persecution of were recorded in crowned eagles’ diet only at those crowned eagles stems from a social and cultural atti- nest sites located close to the (Fig. 1), a tude towards large predators in general, irrespective of habitat feature that was not considered in our coarse, whether predation by raptors or carnivores causes any 2-level habitat categorization. The best models for economic losses (Sarasola & Maceda 2006). In addition, these prey groups included only a constant (Table 2), crowned eagles appear to be flexible dietary general- indicating that the occurrence of these prey was also ists that are not dependent on livestock farming. The independent of the methodological approach em- proportion of some prey items, particularly fish and ployed in the assessment of the diet of crowned reptiles, varied by region. Reptiles were more fre- eagles. The occurrence of reptiles in the diet was also quently recorded at breeding sites located in Loventué independent of habitat type and year but not for the county, and fish were recorded only at nest sites close method employed in the analysis of diet. Reptiles to the Salado River (Fig. 1). The occurrence of fish in were recorded in higher numbers at sites where the the diet was linked to periods in which the river level method employed was direct observation or video dropped as a result of river flow regulation upstream in recording (mean ± SE = 19.2 ± 5.6 reptiles, N = 4 (Fig. 1), and fish either died or were sites) than in those where the diet was examined by confined to small ponds. Crowned eagles might be the analysis of pellets and prey remains (mean ± SE = expected to modify their diet in similar ways when 4.6 ± 2.2 reptiles, N = 9 sites). faced with environmental changes in land use and Sarasola et al.: Feeding ecology and persecution of crowned eagles 211

Table 1. Prey consumed by crowned eagles, determined by analysis of pellets and prey remains, direct observations, and video recording at 15 breeding territories from 2002 to 2009. Results are expressed as absolute frequency for each of the methodologies employed and as total percentage in number (%N) and in mass (%M). Totals are given in bold; (–): 0

Prey Mass Pellets/ Direct Video %N %M (g) prey remains observations records

Mammals 67.73 87.19 Edentata 48.66 70.18 Dwarf Zaedyus pichiy 1020 252 3 12 44.65 57.47 Big hairy armadillo Chaetophractus villosus 3420 17 – – 2.84 12.27 Screaming hairy armadillo Chaetophractus vellerosus 1000 2 – – 0.33 0.42 Unidentified armadillos – – 5 – 0.84 – Carnivora 5.02 13.58 Andes skunk Conepatus chinga 2240 22 – 3 4.18 11.82 Lesser grison cuja 1670 5 – – 0.84 1.76 Rodentia 13.38 2.46 Vesper mice Calomys spp. 16 1 – – 0.17 <0.01 South American grass mice Akodon spp. 25 1 – – 0.17 0.01 Hairy-soled conyrat Reithrodon auritus 78 1 – – 0.17 0.02 Gerbil mice Eligmodontia sp. 20 6 – – 1.00 0.03 -eared mice Phyllotis sp. 57 1 – – 0.17 0.01 Tuco-tuco Ctenomys sp. 120 13 1 – 2.34 0.35 Common yellow-toothed cavy musteloides 225 3 – – 0.50 0.14 Patagonian Dolichotis patagonum 3000a 3 – – 0.50 1.90 Unidentified rodents – 2 4 4 1.67 – Lagomorpha 0.17 0.42 European Lepus europaeus 2000 1 – – 0.17 0.42 Marsupialia 0.33 0.01 Fat-tailed mouse opossum Thylamys sp. 18 2 – – 0.33 0.01 Artiodactyla 0.17 0.53 Goat Capra hircus 2500a 1 – – 0.17 0.53 Birds 3.34 4.49 Monk parakeet Myiopsitta monachus 124 2 – – 0.33 0.05 Greater Rhea americana 3000a 7 – – 1.17 4.43 Tinamous (Tinamidae) – 4 – – 0.67 – Unidentified birds – 7 – – 1.17 – Reptiles 16.22 5.77 Tegus Tupinambis sp. 4500 5 – – 0.84 4.75 Crossed pit viper alternatus 1250 1 – 1 0.33 0.53 Patagonian pit viper Bothrops ammodytoides 1250 1 – – 0.17 0.26 South American hognose snake Lystrophis dorbignyi 300 1 – – 0.17 0.06 Jan’s hognose snake Lystrophis histricus 300 1 – – 0.17 0.06 Green racer snake Philodryas sp. 500 1 – 0.17 0.11 Colubrids – 33 4 23 10.03 – Unidentified reptiles – 2 – – 0.33 – Unidentified snakes – – 23 1 4.01 – Fish 2.51 2.53 Common carp Cyprinus carpio 1000 15 – – 2.51 2.53 15.89 0.02 Carabidae 1.3 44 – – 7.36 0.01 Coleoptera 1.2 30 – – 5.02 0.01 2 3 – – 0.50 <0.01 Scarabidae 1.1 8 – – 1.34 <0.01 Curculionidae 1 8 – – 1.34 <0.01 Unidentified insects – 1 1 – 0.33 – Arachnida 1.00 <0.01 Aranae 1 4 – – 0.66 <0.01 Scorpions 8 2 – – 0.33 <0.01 Total prey items 513 41 44 aEstimated mass for juvenile individuals 212 Endang Species Res 11: 207–213, 2010

Table 2. Best models explaining the consumption of the 4 vertebrate prey surprising under changes in livestock groups by crowned eagles in western La Pampa, Argentina. The intercept and densities. Thus, the current attitude Δ coefficient for the variables included in the top-ranking models ( AICc < 4) for toward crowned eagles could be the each prey group are given with SE in parentheses. For each model, the corrected Akaike’s Information Criterion (AICc), the difference in AICc between result of eagles’ predatory behavior in the current model and the best model (ΔAICc), and the Akaike weights (w) the past and historic perceptions of its are given; where no model name is given, the best model was a null model economic impact. Illegal killing of crowned eagles is Prey group Model Intercept Coefficient AICc ΔAICc w also suspected to occur in some areas of Paraguay (BirdLife International Mammals 3.18 (0.05) 78.53 0.00 0.93 2008) and was probably the main Birds 0.51 (0.22) 50.11 0.00 0.89 cause of local in Uruguay Reptiles METHOD 2.45 (0.10) –0.53 (0.12) 62.83 0.00 0.72 0.38 (0.07) where the last record for a crowned Fish 0.22 (0.25) 55.15 0.00 1.00 eagle came from a shot individual (Alvarez 1933). However, no quantita- tive data exist on direct persecution in agricultural practices. The sheep industry, which those countries or in other parts of the species’ range. played a major role in European settlement after the Given the data presented in this study, further studies eradication of indigenous people in this region, peaked to evaluate the extent of crowned eagle predation on around 1960 (3.5 million sheep for La Pampa province; livestock should be conducted in areas where human DNDEC 1960) but decreased steadily thereafter. The persecution is suspected or documented. reduction in quantity and quality of forage as well as The central and western parts of La Pampa province the use of synthetic materials and other economic fac- have been identified as high-priority areas for field tors were blamed for this decline (Soriano 1983). From research and conservation for crowned eagles, due to 1988 to 2002, for example, the number of sheep in La the high number of field records and to the absence of Pampa province declined from 472 000 to 205 000 other protected areas and reserves (Bellocq et al. head; numbers increased slightly over the same 2002). All of the current known nesting areas of period of time (INDEC 1988, 2002; Fig. 2). By contrast, crowned eagles are managed by private owners, and compose only a small fraction of total livestock any conservation measure will depend strongly on raised in this province (35 000 in 2002). Currently, including the local community in management deci- sheep are mainly raised for subsistence farming by sions and conservation action. Additional management local people rather than commercially, and both sheep and conservation measures would also include moni- and goats are raised in small herds and at low densi- toring and reinforcement of legal regulations. In addi- ties. Livestock losses due to predation may be greater tion, and because crowned eagle predation on live- when herds are larger (Dar et al. 2009), and a shift in stock is almost nonexistent, economic compensation to the feeding behavior of crowned eagles would not be farmers at high rates (e.g. more than 100% of the value of livestock) could be implemented in those cases in which livestock losses due to crowned eagle predation are well documented. This study shows that there is a lack of systematic livestock predation by crowned eagles in central Argentina and suggests that persecution of these birds can be reduced without negatively affecting local agri- cultural practices. The establishment of educational campaigns to disseminate accurate information on eagle diet, rather than costly and difficult actions to mitigate human–wildlife conflicts, are the most urgent and highly prioritized conservation measures that should be taken to preserve crowned eagle popula- tions in this broad region.

Fig. 2. Yearly variation in the total number of cattle and sheep in La Pampa province, Argentina, between 1988 and 2002. No Acknowledgements. We thank C. Solaro, M. Reyes, M. De data are available for sheep in 1998. Data from 1991 to 2000 Benito, L. Pahl, M. Peters, C. Staveley, J. Carslile, M. del Mar are from Encuesta Nacional Agropecuaria (www.indec. Contaldi, and J.J. Maceda for help with field research, and V. mecon.ar) and for 1988 and 2002 from Censo Nacional Salvador for help with data analysis. We thank many local Agropecuario (INDEC 1988, 2002) landowners from La Pampa province who gave us permission Sarasola et al.: Feeding ecology and persecution of crowned eagles 213

to access their ranches and provided information on the loca- Fernandez OA, Busso CA (1997) Arid and semi-arid range- tion of crowned eagle nests and perch sites. We thank L. lands: two thirds of Argentina. Rala Rep 200:41–60 Sympson and the University of California at Berkeley for pro- INDEC (Instituto Nacional de Estadísticas y Censos) (1988) viding the video recording equipment and H. Vargas for pro- Censo nacional agropecuario 1988. Resultados generales viding helpful comments on an early version of this manu- provincia de La Pampa. INDEC, Secretaría de Planifi- script. We appreciate the improvements in English usage cación, made by C. Riehl through the Association of Field Ornitholo- INDEC (2002) Censo nacional agropecuario 2002. Resultados gists’ program of editorial assistance. We also appreciate the generales provincia de La Pampa. www.indec.mecon.ar suggestions and comments made by M. Reed, R. Brook, and López-Lanús B, Grilli P, Di Giacomo AS, Coconier E, Banchs R an anonymous reviewer that greatly improved the manu- (2008) Categorización de las aves de la Argentina según script. This study received financial support from the Pere- su estado de conservación. Aves /AOP y Secre- grine Fund (USA), the Universidad Nacional de La Pampa taría de Ambiente y Desarrollo Sustentable, Buenos Aires (Argentina), IDEA WILD (USA) and the Dirección de Recur- Maceda JJ (2007) Biología y conservación del águila coron- sos Naturales of La Pampa province (Argentina). J.H.S. was ada (Harpyhaliaetus coronatus) en Argentina. Hornero 22: founded by the Spanish Ministry of Education and Science 159–171 through the postdoctoral programme ‘Juan de la Cierva.’ Maceda JJ, Sarasola JH, Pessino MEM (2003) Presas consum- M.A.G. was supported by . idas por el águila coronada (Harpyhaliaetus coronatus) en la provincia de La Pampa. Ornitol Neotrop 14:419–422 Margalida A, Mañosa S, Bertran J, García D (2007) Biases in LITERATURE CITED studying the diet of the bearded vulture. J Wildl Manag 71:1621–1625 Alvarez T (1933) Observaciones biológicas sobre las aves del Matchett MR, O’Gara BW (1987) Methods of controlling Uruguay. An Mus Hist Nat Montev 2:1–50 depredation on domestic sheep in south- Amadon D (1982) A revision of the sub-buteonine hawks. Am western Montana. J Raptor Res 21:85–94 Mus Novit 2741:1–20 MathSoft (1999) S-Plus 2000 guide to statistics. Data Analysis Bellocq MI, Bonaventura SM, Marcelino FN, Sabatini M Products Division, Seattle, WA (1998) Habitat use of crowned eagles (Harpyhaliaetus McCullagh P, Nelder JA (1989) Generalized linear models, coronatus) in the southern limits of the species’ range. J 2nd edn. Chapman & Hall, London Raptor Res 32:312–314 Pearson O (1995) Annotated keys for identifying small mam- Bellocq MI, Ramírez-Llorens P, Filloy J (2002) Recent records mals living in or near Nahuel Huapi National Park or of crowned eagles (Harpyhaliaetus coronatus) from Lanín National Park, southern Argentina. Mastozool Argentina, 1981–2000. J Raptor Res 36:206–212 Neotrop 2:99–148 BirdLife International (2008) Harpyhaliaetus coronatus. In: Redford KH, Eisenberg JF (1992) Mammals of the Neotropics. IUCN Red List of threatened species. Version 2009.1. The southern cone, Vol 2. University of Chicago Press, www.iucnredlist.org. Chicago, IL Burnham KP, Anderson DR (2002) Model selection and multi- Redpath S (1991) The impacts of hen harriers on red grouse model inference: a practical information-theoretic breeding success. J Appl Ecol 28:659–671 approach. Springer-Verlag, New York, NY Sarasola JH, Maceda JJ (2006) Past and current evidence of Cabrera AL (1976) Regiones fitogeográficas Argentinas. In: persecution of the endangered crowned eagle Harpyhali- Ferreira Sobral EF (ed) Enciclopedia Argentina de agricul- aetus coronatus in Argentina. Oryx 40:347–350 tura y jardinería. Editorial ACME, Buenos Aires, p 1–85 Sarasola JH, Santillan MA, Galmes MA (2003) Food habits Cei JM (1986) Reptiles del centro, centro-oeste y sur de la and foraging ecology of American kestrels in the semiarid Argentina. Herpetofauna de zonas áridas y semiáridas. forests of Central Argentina. J Raptor Res 37:236–243 Monografie IV. Museo Regionale di Scienze Naturali, Sarasola JH, Santillan MA, Galmes MA (2007) Comparison Torino, p 112–120 of food habits and prey selection of the white-tailed kite Chebez JC, Maceda JJ, Pereyra- R (2008) Águila coron- (Elanus leucurus) between two land use types in central ada. In: Chebez JC (ed) Los que se van. Fauna Argentina Argentina. Stud Neotrop Fauna Environ 42:85–91 amenazada. Editorial Albatros, Buenos Aires, p 177–186 Soriano A (1983) and semi-deserts of Patagonia. Dar NI, Minhas RA, Zaman Q, Linkie M (2009) Predicting the In: West NE (ed) Ecosystems of the world — temperate patterns, perceptions and causes of human–carnivore deserts and semi-deserts. Elsevier Scientific, Amsterdam, conflict in and around Machiara National Park, Pakistan. p 423–460 Biol Conserv 142:2076–2082 Thirgood S, Redpath S, Newton I, Hudson P (2000) Raptors Davies RAG (1999) The extent, cost and control of livestock and red grouse: conservation conflicts and management predation by eagles with a case study on black eagles solutions. Conserv Biol 14:95–104 (Aquila verreauxii) in the Karoo. J Raptor Res 33:67–72 Thirgood S, Woodroffe R, Rabinowitz A (2005) The impact of DNDEC (Dirección Nacional de Estadísticas y Censos) (1960) human-wildlife conflicts on human lives and livelihoods. Censo nacional agropecuario 1960. DNDEC, Buenos Aires In: Woodroffe S, Thirgood S, Rabinowitz A (eds) People Etheridge B, Summers RW, Green RE (1997) The effects of and wildlife: conflict or coexistence? Cambridge Univer- illegal killing and destruction of nests by humans on the sity Press, Cambridge, p 13–26 population dynamics of the hen harrier Circus cyaneus in Whitfield DP, Fielding AH, McLeod DRA, Haworth PF (2004) Scotland. J Appl Ecol 34:1081–1105 The effects of persecution on age of breeding and territory Fergusson-Lees J, Christie DA (2001) Raptors of the world. occupation in golden eagles in Scotland. Biol Conserv Helm Identification Guides, London 118:249–259

Editorial responsibility: Michael Reed, Submitted: January 5, 2010; Accepted: March 10, 2010 Medford, Massachusetts, USA Proofs received from author(s): April 30, 2010